Concept drift is a term from the fields of artificial intelligence, big data, and digital transformation. It describes a change in the data or patterns recognised by a computer model. This means that data with which a machine or software was once trained changes over time \u2013 it \u201edrifts\u201c. <\/p>\n
Imagine you are using artificial intelligence that detects credit card fraud based on past transaction data. If fraudsters\u201e behaviour changes, the old data no longer matches new fraud attempts. The artificial intelligence eventually recognises novel fraud types less well \u2013 its knowledge therefore becomes \u201coutdated\". This process is called concept drift. <\/p>\n
It is therefore important for companies to regularly review their data models and retrain them with fresh, up-to-date data. Only in this way can a system for fraud detection, prediction of product trends, or analysis of customer behaviour, for example, remain reliable. Understanding concept drift is essential today, in the age of artificial intelligence, to deploy digital systems efficiently and securely.<\/p>","protected":false},"excerpt":{"rendered":"